Identifying Early Cognitive Impairment in Cirrhosis
We compared five new computerized tests to two currently available tests (the number connection test A (NCT-A) and the Inhibitory Control Test (ICT)) with respect to their ability to distinguish the cognitive function of patients with cirrhosis without OHE to that of the following two groups: (a) Pre-cirrhotic chronic liver disease, defined as patients with chronic hepatitis C virus (HCV) infection or NAFLD/NASH (non-alcoholic fatty liver disease/non-alcoholic steatohepatitis) with a biopsy over the preceding 2 years confirming absence of cirrhosis and (b) normal controls with no known liver disease.
Patients with cirrhosis (diagnosed by typical clinical, laboratory, endoscopic and radiological features, or by liver biopsy) or pre-cirrhotic chronic liver disease were recruited between 2010 and 2012 from the Gastroenterology and Hepatology clinics of the Veterans Affairs Puget Sound Health Care System (VAPSHCS) and the University of Washington Medical Center in Seattle, Washington. Normal healthy controls were recruited through community advertising.
Subjects were excluded if they ever had a diagnosis of OHE, if they ever received treatment for hepatic encephalopathy, if they had asterixis or if they scored <25 on the Mini-Mental State Examination (MMSE). Additionally, patients were excluded if they had: conditions that might affect their test performance (dementia, bipolar disorder, manic depression, schizophrenia, psychotic depression, and epilepsy); use of psychoactive medications (antiepileptics, stimulants, sedatives, anxiolytics, hypnotics, and antipsychotics); previous or current treatment with interferon; use of street drugs within the past 3 months; or consumption of alcoholic beverages more than three times per week in the past 3 months or any alcohol in the past 12 h.
Participants completed the MMSE and questionnaires assessing past medical history, social history, alcohol and drug use, and education. Medical records were abstracted for additional information about liver disease, laboratory tests, Child–Turcotte–Pugh Score, and Model for End-Stage Liver Disease (MELD) Score. Liver biopsy slides were reviewed to confirm the stage of liver disease.
We administered the Patient Reported Outcomes Measurement Information System (PROMIS), which comprises a series of computer-adaptive questionnaires that measure specific domains of patient-reported health status and health-related quality of life, validated in cirrhotic patients.
Number Connection Test-A (NCT-A). This paper-and-pencil test involves connecting randomly distributed numbers 1–25 in appropriate sequential order. Subjects completed one trial run and two time-recorded runs using three different NCT-As. The NCT-A is the most commonly used test for encephalopathy in clinical practice.
Inhibitory Control Test (ICT). The ICT is a computerized psychometric test. Different letters appear on the computer screen one at a time every 0.5 s, and the test-taker must hit the spacebar in response to "targets" (any "X" appearing after a "Y" or any "Y" appearing after an "X") while avoiding responding to "lures" (any "X" appearing after an "X" or any "Y" appearing after a "Y"). This test has a "built in" trial run followed by six recorded runs that include a total of 212 "targets" and 40 "lures" interspersed between 1728 random letters. Total administration time is 15 min. The ICT measures attention and response inhibition. It has been validated against existing psychometric tests in cirrhotic patients and has been shown to predict poor driving simulator performance, driving accidents, and development of OHE.
These tests are marketed as "brain-training" games by Lumos Labs (San Francisco, CA). (Figure 1 and see Supplemental Figure S2 online) We downloaded them as an iPhone app called "Lumosity Brain Trainer" in 2010 and administered them on an iPad. We chose these tests because they are computer-based adaptations of pre-existing, non-computerized, validated psychometric tasks that test cognitive domains thought to be important in MHE. Subjects completed one trial run and two test runs for each test.
(Enlarge Image)
Figure 1.
Five brain-training games available as an iPad "app" from Lumosity. (a) "Circles", a test of spatial orientation, information processing speed, and attention. Colored circles appear one at a time and the test-taker must decide whether each is a "Match" or "No match" in terms of color pattern when compared with the previous circle. (b) "Color Match", a version of the Stroop test that evaluates selective attention, cognitive flexibility, and processing speed. The names of two colors appear and the test-taker must read the top word and decide whether or not it matches the font color of the bottom word. (c) "Memory Matrix", a test of visuospatial memory. A pattern of tiles appears on a grid and then disappears, and the test-taker must recreate the pattern by memory by tapping on the correct tiles. (d) "Lost in Migration", a test of attention as well as visual field and focus. Five birds appear on the screen and the test-taker must determine the direction of flight of the bird in the center of the formation. (e) "Chalkboard Challenge", a test of quantitative reasoning. The test-taker must determine which arithmetic figure has the greatest value between two choices. (Images used with permission from Lumos Labs, San Francisco, CA)
(a) "Circles". Colored circles appear one at a time and the test-taker must decide whether each is a "Match" or "No match" in terms of color pattern compared with the previous circle. It is a computer-based adaptation of the N-back (N=1) task, which assesses working memory and speed of processing. For "Circles", as well as for "Color Match" and "Lost in Migration", each run lasts 45 s and the test-taker can give greater or fewer responses depending upon how fast they answer. The total score reflects the total number of correct answers (speed) as well as the number of consecutive correct answers (precision).
(b) "Color Match". The names of two colors appear and the test-taker must read the top word and decide whether or not it matches the font color of the bottom word. It is a computer-based adaptation of the "Stroop task", which measures selective attention, cognitive flexibility, and response inhibition.
(c) "Lost in Migration". Five birds appear on the screen and the test-taker must determine the direction of flight of the bird in the center of the formation. It is a computer-based adaptation of the Eriksen Flanken Task, which assesses visual attention and response inhibition.
(d) "Memory Matrix". A pattern of tiles appears on a grid and then disappears, and the test-taker must recreate the pattern by memory. Difficulty level (i.e., number of tiles) increases or decreases with each subsequent pattern depending on whether the previous answer was correct or incorrect. The game finishes after 15 patterns are completed. Memory matrix is a computer-based adaptation of the Corsi block tapping task, which assesses visuospatial working memory.
(e) "Chalkboard Challenge". The test-taker determines which arithmetic figure has the greatest value between two choices. Correct answers increase the allotted time and difficulty of the questions (and therefore increase the score) while incorrect answers shorten the allotted time, hastening the end of the test. On average each run took 75 s. This test measures executive functions via problem solving and quantitative reasoning.
The tests were administered in the following order: NCT, Circles, Memory Matrix, (SHORT BREAK), Lost in Migration, Color Match, Chalkboard Challenge, (SHORT BREAK), ICT.
For the NCT and the five "brain-training games", the average of the two test runs was used for analyses. The ICT has six built-in test runs yielding a total score. We compared cirrhotic patients with each of the two comparison groups with respect to the mean scores obtained in each of the tests using linear regression, with or without adjusting for age and educational attainment.
We compared the scores of the two test runs that each participant performed for each of the five "brain-training games" and the NCT using the Pearson's r correlation to determine their test–retest reliability. The ICT was administered only once due its long administration time; hence test–retest reliability could not be measured.
We plotted receiver operating characteristic (ROC) curves to determine how well each test distinguished patients with cirrhosis from patients with pre-cirrhotic liver disease and calculated the area under the curve (and 95% confidence interval (CI)).
We used multivariate logistic regression to determine whether combinations of "brain-training games" discriminated between cirrhosis and pre-cirrhotic liver disease after adjusting for age and education.
The study was approved by the Institutional Review Boards of VAPSHCS and the University of Washington.
Methods
Study Design
We compared five new computerized tests to two currently available tests (the number connection test A (NCT-A) and the Inhibitory Control Test (ICT)) with respect to their ability to distinguish the cognitive function of patients with cirrhosis without OHE to that of the following two groups: (a) Pre-cirrhotic chronic liver disease, defined as patients with chronic hepatitis C virus (HCV) infection or NAFLD/NASH (non-alcoholic fatty liver disease/non-alcoholic steatohepatitis) with a biopsy over the preceding 2 years confirming absence of cirrhosis and (b) normal controls with no known liver disease.
Study Participants
Patients with cirrhosis (diagnosed by typical clinical, laboratory, endoscopic and radiological features, or by liver biopsy) or pre-cirrhotic chronic liver disease were recruited between 2010 and 2012 from the Gastroenterology and Hepatology clinics of the Veterans Affairs Puget Sound Health Care System (VAPSHCS) and the University of Washington Medical Center in Seattle, Washington. Normal healthy controls were recruited through community advertising.
Subjects were excluded if they ever had a diagnosis of OHE, if they ever received treatment for hepatic encephalopathy, if they had asterixis or if they scored <25 on the Mini-Mental State Examination (MMSE). Additionally, patients were excluded if they had: conditions that might affect their test performance (dementia, bipolar disorder, manic depression, schizophrenia, psychotic depression, and epilepsy); use of psychoactive medications (antiepileptics, stimulants, sedatives, anxiolytics, hypnotics, and antipsychotics); previous or current treatment with interferon; use of street drugs within the past 3 months; or consumption of alcoholic beverages more than three times per week in the past 3 months or any alcohol in the past 12 h.
Study Procedures
Participants completed the MMSE and questionnaires assessing past medical history, social history, alcohol and drug use, and education. Medical records were abstracted for additional information about liver disease, laboratory tests, Child–Turcotte–Pugh Score, and Model for End-Stage Liver Disease (MELD) Score. Liver biopsy slides were reviewed to confirm the stage of liver disease.
We administered the Patient Reported Outcomes Measurement Information System (PROMIS), which comprises a series of computer-adaptive questionnaires that measure specific domains of patient-reported health status and health-related quality of life, validated in cirrhotic patients.
Tests of Cognitive Impairment
Number Connection Test-A (NCT-A). This paper-and-pencil test involves connecting randomly distributed numbers 1–25 in appropriate sequential order. Subjects completed one trial run and two time-recorded runs using three different NCT-As. The NCT-A is the most commonly used test for encephalopathy in clinical practice.
Inhibitory Control Test (ICT). The ICT is a computerized psychometric test. Different letters appear on the computer screen one at a time every 0.5 s, and the test-taker must hit the spacebar in response to "targets" (any "X" appearing after a "Y" or any "Y" appearing after an "X") while avoiding responding to "lures" (any "X" appearing after an "X" or any "Y" appearing after a "Y"). This test has a "built in" trial run followed by six recorded runs that include a total of 212 "targets" and 40 "lures" interspersed between 1728 random letters. Total administration time is 15 min. The ICT measures attention and response inhibition. It has been validated against existing psychometric tests in cirrhotic patients and has been shown to predict poor driving simulator performance, driving accidents, and development of OHE.
New Computerized Tests for Cognitive Impairment in Cirrhosis
These tests are marketed as "brain-training" games by Lumos Labs (San Francisco, CA). (Figure 1 and see Supplemental Figure S2 online) We downloaded them as an iPhone app called "Lumosity Brain Trainer" in 2010 and administered them on an iPad. We chose these tests because they are computer-based adaptations of pre-existing, non-computerized, validated psychometric tasks that test cognitive domains thought to be important in MHE. Subjects completed one trial run and two test runs for each test.
(Enlarge Image)
Figure 1.
Five brain-training games available as an iPad "app" from Lumosity. (a) "Circles", a test of spatial orientation, information processing speed, and attention. Colored circles appear one at a time and the test-taker must decide whether each is a "Match" or "No match" in terms of color pattern when compared with the previous circle. (b) "Color Match", a version of the Stroop test that evaluates selective attention, cognitive flexibility, and processing speed. The names of two colors appear and the test-taker must read the top word and decide whether or not it matches the font color of the bottom word. (c) "Memory Matrix", a test of visuospatial memory. A pattern of tiles appears on a grid and then disappears, and the test-taker must recreate the pattern by memory by tapping on the correct tiles. (d) "Lost in Migration", a test of attention as well as visual field and focus. Five birds appear on the screen and the test-taker must determine the direction of flight of the bird in the center of the formation. (e) "Chalkboard Challenge", a test of quantitative reasoning. The test-taker must determine which arithmetic figure has the greatest value between two choices. (Images used with permission from Lumos Labs, San Francisco, CA)
(a) "Circles". Colored circles appear one at a time and the test-taker must decide whether each is a "Match" or "No match" in terms of color pattern compared with the previous circle. It is a computer-based adaptation of the N-back (N=1) task, which assesses working memory and speed of processing. For "Circles", as well as for "Color Match" and "Lost in Migration", each run lasts 45 s and the test-taker can give greater or fewer responses depending upon how fast they answer. The total score reflects the total number of correct answers (speed) as well as the number of consecutive correct answers (precision).
(b) "Color Match". The names of two colors appear and the test-taker must read the top word and decide whether or not it matches the font color of the bottom word. It is a computer-based adaptation of the "Stroop task", which measures selective attention, cognitive flexibility, and response inhibition.
(c) "Lost in Migration". Five birds appear on the screen and the test-taker must determine the direction of flight of the bird in the center of the formation. It is a computer-based adaptation of the Eriksen Flanken Task, which assesses visual attention and response inhibition.
(d) "Memory Matrix". A pattern of tiles appears on a grid and then disappears, and the test-taker must recreate the pattern by memory. Difficulty level (i.e., number of tiles) increases or decreases with each subsequent pattern depending on whether the previous answer was correct or incorrect. The game finishes after 15 patterns are completed. Memory matrix is a computer-based adaptation of the Corsi block tapping task, which assesses visuospatial working memory.
(e) "Chalkboard Challenge". The test-taker determines which arithmetic figure has the greatest value between two choices. Correct answers increase the allotted time and difficulty of the questions (and therefore increase the score) while incorrect answers shorten the allotted time, hastening the end of the test. On average each run took 75 s. This test measures executive functions via problem solving and quantitative reasoning.
The tests were administered in the following order: NCT, Circles, Memory Matrix, (SHORT BREAK), Lost in Migration, Color Match, Chalkboard Challenge, (SHORT BREAK), ICT.
Statistical Analysis
For the NCT and the five "brain-training games", the average of the two test runs was used for analyses. The ICT has six built-in test runs yielding a total score. We compared cirrhotic patients with each of the two comparison groups with respect to the mean scores obtained in each of the tests using linear regression, with or without adjusting for age and educational attainment.
We compared the scores of the two test runs that each participant performed for each of the five "brain-training games" and the NCT using the Pearson's r correlation to determine their test–retest reliability. The ICT was administered only once due its long administration time; hence test–retest reliability could not be measured.
We plotted receiver operating characteristic (ROC) curves to determine how well each test distinguished patients with cirrhosis from patients with pre-cirrhotic liver disease and calculated the area under the curve (and 95% confidence interval (CI)).
We used multivariate logistic regression to determine whether combinations of "brain-training games" discriminated between cirrhosis and pre-cirrhotic liver disease after adjusting for age and education.
The study was approved by the Institutional Review Boards of VAPSHCS and the University of Washington.
SHARE
